---

Machine Learning Model Evaluation

This repository contains a collection of machine learning models and their respective performance evaluations on a specific dataset. Each model is trained and tested on the dataset, and their accuracy scores are reported for reference.

Models and Their Performance

1. Stochastic Gradient Descent (SGD)

• Accuracy Score: 57.69%

The Stochastic Gradient Descent model achieved an accuracy score of 57.69%. This score reflects the model's ability to make correct predictions on the given dataset.

2. Linear Support Vector Classifier (SVC)

• Accuracy Score: 50.0%

The Linear Support Vector Classifier yielded an accuracy score of 50.0%. This score measures the model's classification accuracy on the dataset.

3. SVC Parameter Tuning with GridSearch

• Accuracy: 53.85%

By applying hyperparameter tuning with GridSearch, the Support Vector Classifier improved its accuracy to 53.85%. This demonstrates the impact of hyperparameter optimization on model performance.

4. K-Nearest Neighbors (KNN)

• Accuracy: 76.9%

The K-Nearest Neighbors model achieved a notable accuracy score of 76.9%. This high accuracy suggests that KNN is well-suited for the given dataset.

5. Decision Tree

• Accuracy: 53.85%

The Decision Tree model achieved an accuracy score of 53.85%. This score reflects the model's performance in making decisions based on the dataset's features.

6. Logistic Regression

• Accuracy: 62.0%

Logistic Regression achieved an accuracy score of 62.0%. This logistic regression model is suitable for binary classification tasks and performed well on this dataset.

7. Random Forest

• Accuracy Score: 73.6%

The Random Forest model achieved an accuracy score of 73.6%. Random Forest is an ensemble model, and its high accuracy reflects its ability to combine multiple decision trees effectively.

Hyperparameter Tuning and Model Details

In addition to presenting model accuracy scores, this repository includes information about hyperparameter tuning and model details:

• Hyperparameter tuning using RandomizedSearch CV: The Random Forest model was fine-tuned using RandomizedSearch Cross-Validation to optimize its hyperparameters.

• Hyperparameter Tuning - GridSearchCV: The Support Vector Classifier was enhanced through GridSearchCV, allowing us to find the best combination of hyperparameters for improved performance.

• Random Forest Classifier model with default parameters: The initial performance of the Random Forest model using default parameters is provided for reference.

• Feature Scaling: Note that feature scaling techniques were applied to ensure that the input features were properly scaled before model training.

---